-
Notifications
You must be signed in to change notification settings - Fork 6
Research Hub Overview
The Research Hub Framework is a framework that scaffolds and guides Data Hub users and developers in building out a multi-entity discovery application with a web GUI and a semantic graph.
From the Extension and Coding Guide (attached below):
- A Framework to allow a combined team of business analysts and developers to create Data Hubs, together with a Search and Discovery application to expose the data
- Data Integration
- Integrates many data sets together – like any MarkLogic Data Hub
- Supports structured and unstructured (or semi-structured) data sources
- Combines an appropriate subset of the data into a knowledge graph using RDF and semantic technology
- Masters data where appropriate (in the MDM sense)
- Provides APIs for search, discovery, updates, reporting, business intelligence
- Search, Discovery and Knowledge Sharing GUI
- Search for any type of data
- See the data as results and details pages
- See a graph view to provide context around the data and discover new items
- Support users gathering and saving items into Workspaces for sharing and Knowledge Transfer
- Support users gathering and saving items into Workspaces for long-running discovery processes
- Search for any type of data
- AI
- Recommend relevant content based on current and past activities
- Adding to Workspaces guides search results, by default
- Including items on the graph view guides graph-based Recommendations (using link walking and analysis)
- Extensible to any feature set that MarkLogic generally supports, including compliance, 360-degree views of key business entities, analytics, powering other applications in an enterprise via REST or SOAP calls
A lot of the capabilities in the framework, and steps in the document, are regular "Data Hub" implementation steps. Building flows and steps within those flows, implementing model-to-model data mappings, inserting triples via TDEs (template driven extraction templates). But some of it is specific guidance that comprise a tried-and-true path to building a powerful data hub, and of course the framework provides code and examples to accelerate along this path.
Also, because there is a complex "pharma-research" example included, and a much simpler HR360 example, teams can start by extending and tweaking, rather than trying to invent a Data Hub application architecture from scratch. For new users of the Research Hub (or the Data Hub) a good place to start is the simpler HR360 example.
If HR and pharma data is not what you need, or the specific data sets are not ideal for you, start with an "empty" project and build your own entities. If those data sets work for you, you might start with one of the example applications and modify it.
Getting Started Check out more documentation and videos here: Research Hub Configuration and Coding Guidance